即使在引发异常的情况下，继续在Python中的try子句中

Question

Is there any way to continue code execution in a try clause, even when the exception is raised?

This is the code example.

outer_margin = 0.125  # Default Amazon
safe_zone = 0.25  # Default Amazon
safe_spine = 0.0625  # Default Amazon
if custom_measure is not None:
    if isinstance(custom_measure, dict):
        try:
            outer_margin = custom_measure["outer_margin"]
            safe_zone = custom_measure["safe_zone"]
            safe_spine = custom_measure["safe_spine"]
        except KeyError as e:
            logger.warning("Key {} not found, default value used.".format(e))

All three values are optional, so the key outer_margin may not exist, but safe_zone may be present. I would like to reach safe_zone even if calling outer_margin raises an exception. Is there any way to do this in a single try/except block? Something like:

# Not valid syntax
try:
    outer_margin = custom_measure["outer_margin"]
try:
    safe_zone = custom_measure["safe_zone"]
except KeyError as e:
    pass

I already found a different solution to the case in the example, but now I am curious to know if there is a way to continue down one try/except block, even if the exception is raised.

Answer 1

You can use dict.get instead of catching the KeyError exception.

if isinstance(custom_measure, dict):
    outer_margin = custom_measure.get("outer_margin")
    safe_zone = custom_measure.get("safe_zone")
    safe_spine = custom_measure.get("safe_spine")

if "outer_margin" does not exist in custom_measure it will be set to None .

You can also set another default value than None like this.

outer_margin = custom_measure.get("outer_margin", 0)

Answer 2

If you really need the log, refactor the try/except out of there:

def get_with_warning(data, key, default):
    if key in data:
        return data[key]
    else:
        logger.warning("Key {} not found, default value used.".format(key))
        return default

outer_margin = get_with_warning(custom_measure, "outer_margin", 0.125)
safe_zone = get_with_warning(custom_measure, "safe_zone", 0.25)
safe_spine = get_with_warning(custom_measure, "safe_spine", 0.0625)

If not, then it's as easy as this:

outer_margin = custom_measure.get("outer_margin", 0.125)
safe_zone = custom_measure.get("safe_zone", 0.25)
safe_spine = custom_measure.get("safe_spine", 0.0625)